Instead of having a global lock for potentially less contention.
Signed-off-by: Christian König <christian.koenig@amd.com>
Tested-by: Nirmoy Das <nirmoy.das@amd.com>
Reviewed-by: Huang Rui <ray.huang@amd.com>
Reviewed-by: Matthew Auld <matthew.auld@intel.com>
Link: https://patchwork.freedesktop.org/patch/424010/
Instead evict round robin from each devices SYSTEM and TT domain.
v2: reorder num_pages access reported by Dan's script
v3: fix rebase fallout, num_pages should be 32bit
Signed-off-by: Christian König <christian.koenig@amd.com>
Tested-by: Nirmoy Das <nirmoy.das@amd.com>
Reviewed-by: Huang Rui <ray.huang@amd.com>
Reviewed-by: Matthew Auld <matthew.auld@intel.com>
Link: https://patchwork.freedesktop.org/patch/424009/
Move the iteration of the global lru into the new function
ttm_global_swapout() and use that instead in drivers.
v2: consistently return int
v3: fix build fail
Signed-off-by: Christian König <christian.koenig@amd.com>
Tested-by: Nirmoy Das <nirmoy.das@amd.com>
Reviewed-by: Huang Rui <ray.huang@amd.com>
Reviewed-by: Matthew Auld <matthew.auld@intel.com>
Link: https://patchwork.freedesktop.org/patch/424008/
QXL indeed unrefs pinned BOs and the warnings are spamming peoples log files.
Make sure we warn only once until the QXL driver is fixed.
Signed-off-by: Christian König <christian.koenig@amd.com>
References: https://lore.kernel.org/lkml/YD+eYcMMcdlXB8PY@alley/
Link: https://patchwork.freedesktop.org/patch/422834/
Reviewed-by: Daniel Vetter <daniel@ffwll.ch>
Signed-off-by: Maarten Lankhorst <maarten.lankhorst@linux.intel.com>
BO would be added into swap list if it is validated into system domain.
If BO is validated again into non-system domain, say, VRAM domain. It
actually should not be in the swap list.
Signed-off-by: xinhui pan <xinhui.pan@amd.com>
Acked-by: Guchun Chen <guchun.chen@amd.com>
Acked-by: Alex Deucher <alexander.deucher@amd.com>
Reviewed-by: Christian König <christian.koenig@amd.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20210224032808.150465-1-xinhui.pan@amd.com
Signed-off-by: Christian König <christian.koenig@amd.com>
This is just another feature which is only used by VMWGFX, so move
it into the driver instead.
I've tried to add the accounting sysfs file to the kobject of the drm
minor, but I'm not 100% sure if this works as expected.
v2: fix typo in KFD and avoid 64bit divide
v3: fix init order in VMWGFX
v4: use pdev sysfs reference instead of drm
Signed-off-by: Christian König <christian.koenig@amd.com>
Reviewed-by: Zack Rusin <zackr@vmware.com> (v3)
Tested-by: Nirmoy Das <nirmoy.das@amd.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20210208133226.36955-2-christian.koenig@amd.com
TTM implements a rather extensive accounting of allocated memory.
There are two reasons for this:
1. It tries to block userspace allocating a huge number of very small
BOs without accounting for the kmalloced memory.
2. Make sure we don't over allocate and run into an OOM situation
during swapout while trying to handle the memory shortage.
This is only partially a good idea. First of all it is perfectly
valid for an application to use all of system memory, limiting it to
50% is not really acceptable.
What we need to take care of is that the application is held
accountable for the memory it allocated. This is what control
mechanisms like memcg and the normal Linux page accounting already do.
Making sure that we don't run into an OOM situation while trying to
cope with a memory shortage is still a good idea, but this is also
not very well implemented since it means another opportunity of
recursion from the driver back into TTM.
So start to rework all of this by implementing a shrinker callback which
allows for TT object to be swapped out if necessary.
v2: Switch from limit to shrinker callback.
v3: fix gfp mask handling, use atomic for swapable_pages, add debugfs
v4: drop the extra gfp_mask checks
Signed-off-by: Christian König <christian.koenig@amd.com>
Reviewed-by: Daniel Vetter <daniel.vetter@ffwll.ch>
Link: https://patchwork.freedesktop.org/patch/msgid/20210208133226.36955-1-christian.koenig@amd.com
Rename ttm_bo_device to ttm_device.
Rename ttm_bo_driver to ttm_device_funcs.
Rename ttm_bo_global to ttm_global.
Move global and device related functions to ttm_device.[ch].
No functional change.
Signed-off-by: Christian König <christian.koenig@amd.com>
Acked-by: Daniel Vetter <daniel.vetter@ffwll.ch>
Link: https://patchwork.freedesktop.org/patch/415222/
As far as I can tell the buffer_count was never used by an
userspace application.
The number of BOs in the system is far better suited in
debugfs than sysfs and we now should be able to add other
information here as well.
v2: add that additionally to sysfs
Signed-off-by: Christian König <christian.koenig@amd.com>
Acked-by: Daniel Vetter <daniel.vetter@ffwll.ch>
Link: https://patchwork.freedesktop.org/patch/414954/
Recently a regression was introduced which caused TTM's buffer eviction to
attempt to evict already-pinned BOs, causing issues with buffer eviction
under memory pressure along with suspend/resume:
nouveau 0000:1f:00.0: DRM: evicting buffers...
nouveau 0000:1f:00.0: DRM: Moving pinned object 00000000c428c3ff!
nouveau 0000:1f:00.0: fifo: fault 00 [READ] at 0000000000200000 engine 04
[BAR1] client 07 [HUB/HOST_CPU] reason 02 [PTE] on channel -1 [00ffeaa000
unknown]
nouveau 0000:1f:00.0: fifo: DROPPED_MMU_FAULT 00001000
nouveau 0000:1f:00.0: fifo: fault 01 [WRITE] at 0000000000020000 engine
0c [HOST6] client 07 [HUB/HOST_CPU] reason 02 [PTE] on channel 1
[00ffb28000 DRM]
nouveau 0000:1f:00.0: fifo: channel 1: killed
nouveau 0000:1f:00.0: fifo: runlist 0: scheduled for recovery
[TTM] Buffer eviction failed
nouveau 0000:1f:00.0: DRM: waiting for kernel channels to go idle...
nouveau 0000:1f:00.0: DRM: failed to idle channel 1 [DRM]
nouveau 0000:1f:00.0: DRM: resuming display...
After some bisection and investigation, it appears this resulted from the
recent changes to ttm_bo_move_to_lru_tail(). Previously when a buffer was
pinned, the buffer would be removed from the LRU once ttm_bo_unreserve
to maintain the LRU list when pinning or unpinning BOs. However, since:
commit 3d1a88e105 ("drm/ttm: cleanup LRU handling further")
We've been exiting from ttm_bo_move_to_lru_tail() at the very beginning of
the function if the bo we're looking at is pinned, resulting in the pinned
BO never getting removed from the lru and as a result - causing issues when
it eventually becomes time for eviction.
So, let's fix this by calling ttm_bo_del_from_lru() from
ttm_bo_move_to_lru_tail() in the event that we're dealing with a pinned
buffer.
v2 (chk): reduce to only the fixing one liner since we always want to
call the callback whenever we would move on the LRU.
Fixes: 3d1a88e105 ("drm/ttm: cleanup LRU handling further")
Cc: Dave Airlie <airlied@redhat.com>
Reviewed-by: Christian König <christian.koenig@amd.com>
Signed-off-by: Lyude Paul <lyude@redhat.com>
Signed-off-by: Christian König <christian.koenig@amd.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20210105114505.38210-1-christian.koenig@amd.com
We only completely delete the BO from the LRU on destruction.
Signed-off-by: Christian König <christian.koenig@amd.com>
Acked-by: Andrey Grodzovsky <andrey.grodzovsky@amd.com>
Link: https://patchwork.freedesktop.org/patch/404618/
Check the pin_count instead of the lru list is empty here.
Signed-off-by: Christian König <christian.koenig@amd.com>
Acked-by: Andrey Grodzovsky <andrey.grodzovsky@amd.com>
Link: https://patchwork.freedesktop.org/patch/404617/
Based on an idea from Dave, but cleaned up a bit.
We had multiple fields for essentially the same thing.
Now bo->base.size is the original size of the BO in
arbitrary units, usually bytes.
bo->mem.num_pages is the size in number of pages in the
resource domain of bo->mem.mem_type.
v2: use the GEM object size instead of the BO size
v3: fix printks in some places
Signed-off-by: Christian König <christian.koenig@amd.com>
Reviewed-by: Huang Rui <ray.huang@amd.com> (v1)
Acked-by: Daniel Vetter <daniel.vetter@ffwll.ch>
Link: https://patchwork.freedesktop.org/patch/406831/
ttm_module.h deals with internals of TTM and should never
be include outside of it.
v2: also move the file around
Signed-off-by: Christian König <christian.koenig@amd.com>
Acked-by: Daniel Vetter <daniel.vetter@ffwll.ch>
Link: https://patchwork.freedesktop.org/patch/404885/
Not technically a problem for ttm, but very likely a driver bug and
pretty big time confusing for reviewing code.
So warn about it, both at cleanup time (so we catch these for sure)
and at pin/unpin time (so we know who's the culprit).
Reviewed-by: Huang Rui <ray.huang@amd.com>
Reviewed-by: Christian König <christian.koenig@amd.com>
Signed-off-by: Daniel Vetter <daniel.vetter@intel.com>
Cc: Christian Koenig <christian.koenig@amd.com>
Cc: Huang Rui <ray.huang@amd.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20201028113120.3641237-1-daniel.vetter@ffwll.ch
Fixes the following W=1 kernel build warning(s):
drivers/gpu/drm/ttm/ttm_bo.c:51: warning: Function parameter or member 'ttm_global_mutex' not described in 'DEFINE_MUTEX'
drivers/gpu/drm/ttm/ttm_bo.c:286: warning: Function parameter or member 'bo' not described in 'ttm_bo_cleanup_memtype_use'
drivers/gpu/drm/ttm/ttm_bo.c:359: warning: Function parameter or member 'bo' not described in 'ttm_bo_cleanup_refs'
drivers/gpu/drm/ttm/ttm_bo.c:359: warning: Function parameter or member 'interruptible' not described in 'ttm_bo_cleanup_refs'
drivers/gpu/drm/ttm/ttm_bo.c:359: warning: Function parameter or member 'no_wait_gpu' not described in 'ttm_bo_cleanup_refs'
drivers/gpu/drm/ttm/ttm_bo.c:359: warning: Function parameter or member 'unlock_resv' not described in 'ttm_bo_cleanup_refs'
drivers/gpu/drm/ttm/ttm_bo.c:424: warning: Function parameter or member 'bdev' not described in 'ttm_bo_delayed_delete'
drivers/gpu/drm/ttm/ttm_bo.c:424: warning: Function parameter or member 'remove_all' not described in 'ttm_bo_delayed_delete'
drivers/gpu/drm/ttm/ttm_bo.c:635: warning: Function parameter or member 'bo' not described in 'ttm_bo_evict_swapout_allowable'
drivers/gpu/drm/ttm/ttm_bo.c:635: warning: Function parameter or member 'ctx' not described in 'ttm_bo_evict_swapout_allowable'
drivers/gpu/drm/ttm/ttm_bo.c:635: warning: Function parameter or member 'locked' not described in 'ttm_bo_evict_swapout_allowable'
drivers/gpu/drm/ttm/ttm_bo.c:635: warning: Function parameter or member 'busy' not described in 'ttm_bo_evict_swapout_allowable'
drivers/gpu/drm/ttm/ttm_bo.c:769: warning: Function parameter or member 'bo' not described in 'ttm_bo_add_move_fence'
drivers/gpu/drm/ttm/ttm_bo.c:769: warning: Function parameter or member 'man' not described in 'ttm_bo_add_move_fence'
drivers/gpu/drm/ttm/ttm_bo.c:769: warning: Function parameter or member 'mem' not described in 'ttm_bo_add_move_fence'
drivers/gpu/drm/ttm/ttm_bo.c:769: warning: Function parameter or member 'no_wait_gpu' not described in 'ttm_bo_add_move_fence'
drivers/gpu/drm/ttm/ttm_bo.c:806: warning: Function parameter or member 'bo' not described in 'ttm_bo_mem_force_space'
drivers/gpu/drm/ttm/ttm_bo.c:806: warning: Function parameter or member 'place' not described in 'ttm_bo_mem_force_space'
drivers/gpu/drm/ttm/ttm_bo.c:806: warning: Function parameter or member 'mem' not described in 'ttm_bo_mem_force_space'
drivers/gpu/drm/ttm/ttm_bo.c:806: warning: Function parameter or member 'ctx' not described in 'ttm_bo_mem_force_space'
drivers/gpu/drm/ttm/ttm_bo.c:872: warning: Function parameter or member 'bo' not described in 'ttm_bo_mem_space'
drivers/gpu/drm/ttm/ttm_bo.c:872: warning: Function parameter or member 'placement' not described in 'ttm_bo_mem_space'
drivers/gpu/drm/ttm/ttm_bo.c:872: warning: Function parameter or member 'mem' not described in 'ttm_bo_mem_space'
drivers/gpu/drm/ttm/ttm_bo.c:872: warning: Function parameter or member 'ctx' not described in 'ttm_bo_mem_space'
drivers/gpu/drm/ttm/ttm_bo.c:1387: warning: Function parameter or member 'ctx' not described in 'ttm_bo_swapout'
Reviewed-by: Christian König <christian.koenig@amd.com>
Cc: Christian Koenig <christian.koenig@amd.com>
Cc: Huang Rui <ray.huang@amd.com>
Cc: David Airlie <airlied@linux.ie>
Cc: Daniel Vetter <daniel@ffwll.ch>
Cc: Sumit Semwal <sumit.semwal@linaro.org>
Cc: dri-devel@lists.freedesktop.org
Cc: linux-media@vger.kernel.org
Cc: linaro-mm-sig@lists.linaro.org
Signed-off-by: Lee Jones <lee.jones@linaro.org>
Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
Link: https://patchwork.freedesktop.org/patch/msgid/20201116174112.1833368-32-lee.jones@linaro.org
Currently drivers get called to move a buffer, but if they have to
move it temporarily through another space (SYSTEM->VRAM via TT)
then they can end up with a lot of ttm->driver->ttm call stacks,
if the temprorary space moves requires eviction.
Instead of letting the driver do all the placement/space for the
temporary, allow it to report back (-EMULTIHOP) and a placement (hop)
to the move code, which will then do the temporary move, and the
correct placement move afterwards.
This removes a lot of code from drivers, at the expense of
adding some midlayering. I've some further ideas on how to turn
it inside out, but I think this is a good solution to the call
stack problems.
v2: separate out the driver patches, add WARN for getting
MULTHOP in paths we shouldn't (Daniel)
v3: use memset (Christian)
Reviewed-by: Daniel Vetter <daniel.vetter@ffwll.ch>
Reviewed-by: hristian König <christian.koenig@amd.com>
Signed-off-by: Dave Airlie <airlied@redhat.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20201109005432.861936-2-airlied@gmail.com
The ttm_operation_ctx structure has a mixture of flags and bools. Drop the
flags and replace them with bools as well.
v2: fix typos, improve comments
Signed-off-by: Christian König <christian.koenig@amd.com>
Reviewed-by: Daniel Vetter <daniel.vetter@ffwll.ch>
Link: https://patchwork.freedesktop.org/patch/398686/
Not used any more.
Signed-off-by: Christian König <christian.koenig@amd.com>
Reviewed-by: Dave Airlie <airlied@redhat.com>
Reviewed-by: Madhav Chauhan <madhav.chauhan@amd.com>
Tested-by: Huang Rui <ray.huang@amd.com>
Link: https://patchwork.freedesktop.org/patch/397087/?series=83051&rev=1
Provide the necessary parameters by all drivers and use the new pool alloc
when no driver specific function is provided.
v2: fix the GEM VRAM helpers
Signed-off-by: Christian König <christian.koenig@amd.com>
Reviewed-by: Dave Airlie <airlied@redhat.com>
Reviewed-by: Madhav Chauhan <madhav.chauhan@amd.com>
Tested-by: Huang Rui <ray.huang@amd.com>
Link: https://patchwork.freedesktop.org/patch/397081/?series=83051&rev=1
It makes no difference to kmalloc if the structure
is 48 or 64 bytes in size.
Signed-off-by: Christian König <christian.koenig@amd.com>
Reviewed-by: Dave Airlie <airlied@redhat.com>
Link: https://patchwork.freedesktop.org/patch/396950/
The move notify callback is only used in one place, this should
be removed in the future, but for now just rename it to the use
case which is to notify the driver that the GPU memory is to be
deleted.
Drivers can be cleaned up after this separately.
Reviewed-by: Christian König <christian.koenig@amd.com>
Signed-off-by: Dave Airlie <airlied@redhat.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20201021044031.1752624-2-airlied@gmail.com
Changing the caching on the fly never really worked
flawlessly.
So stop this completely and just let drivers specific the
desired caching in the tt or bus object.
Signed-off-by: Christian König <christian.koenig@amd.com>
Reviewed-by: Michael J. Ruhl <michael.j.ruhl@intel.com>
Link: https://patchwork.freedesktop.org/patch/394256/
It is the sole user of this.
Signed-off-by: Christian König <christian.koenig@amd.com>
Acked-by: Daniel Vetter <daniel.vetter@ffwll.ch>
Link: https://patchwork.freedesktop.org/patch/393498/
We can always access the global state.
Signed-off-by: Christian König <christian.koenig@amd.com>
Acked-by: Daniel Vetter <daniel.vetter@ffwll.ch>
Link: https://patchwork.freedesktop.org/patch/393499/
Make it more clear what the resource manager function
does and nuke the wrapper function.
v2: nuke the wrapper
v3: fix typo in radeon, rebased
Signed-off-by: Christian König <christian.koenig@amd.com>
Reviewed-by: Daniel Vetter <daniel.vetter@ffwll.ch> (v2)
Link: https://patchwork.freedesktop.org/patch/393914/
As an alternative to the placement flag add a
pin count to the ttm buffer object.
v2: add dma_resv_assert_help() calls
Signed-off-by: Christian König <christian.koenig@amd.com>
Reviewed-by: Dave Airlie <airlied@redhat.com>
Reviewed-by: Huang Rui <ray.huang@amd.com>
Link: https://patchwork.freedesktop.org/patch/391596/?series=81973&rev=1
When we swapout/in a BO we try to change the caching
attributes of the pages before/after doing the copy.
On x86 this is done by calling set_pages_uc(),
set_memory_wc() or set_pages_wb() for not highmem pages
to update the linear mapping of the page.
On all other platforms we do exactly nothing.
Now on x86 this is unnecessary because copy_highpage() will
either create a temporary mapping of the page which is wb
anyway and destroyed immediately again or use the linear
mapping with the correct caching attributes.
So stop this nonsense and just keep the caching as it is and
return an error when a driver tries to change the caching of
an already populated TT object.
This is much more defensive since changing caching
attributes is platform and driver specific and usually
doesn't work after the page was initially allocated.
Signed-off-by: Christian König <christian.koenig@amd.com>
Reviewed-by: Alex Deucher <alexander.deucher@amd.com>
Link: https://patchwork.freedesktop.org/patch/391293/
This moves the generic tracking into the drivers and protects
against reentrancy in the drivers. It fixes up radeon and agp
to be able to query the bound status as that is required.
Reviewed-by: Christian König <christian.koenig@amd.com>
Signed-off-by: Dave Airlie <airlied@redhat.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20200917043040.146575-2-airlied@gmail.com
Unexport ttm_check_under_lowerlimit.
Make ttm_bo_acc_size static and unexport it.
Remove ttm_get_kernel_zone_memory_size.
Signed-off-by: Christian König <christian.koenig@amd.com>
Reviewed-by: Dave Airlie <airlied@redhat.com>
Link: https://patchwork.freedesktop.org/patch/390515/
Instead of letting TTM make an educated guess based on
some mask all drivers should just specify what caching
they want for their CPU mappings.
Signed-off-by: Christian König <christian.koenig@amd.com>
Reviewed-by: Dave Airlie <airlied@redhat.com>
Link: https://patchwork.freedesktop.org/patch/390207/
As far as I can tell this was never used either and we just
always fallback to the order cached > wc > uncached anyway.
Signed-off-by: Christian König <christian.koenig@amd.com>
Reviewed-by: Dave Airlie <airlied@redhat.com>
Link: https://patchwork.freedesktop.org/patch/390142/
It's not supported to specify more than one of those flags.
So it never made sense to make this a flag in the first place.
Nuke the flags and specify directly which memory type to use.
Signed-off-by: Christian König <christian.koenig@amd.com>
Reviewed-by: Dave Airlie <airlied@redhat.com>
Link: https://patchwork.freedesktop.org/patch/389826/?series=81551&rev=1
This is used by TTM to communicate the physical address
which should be used with ioremap(), ioremap_wc(). We don't
need to separate the base and offset in any way here.
Signed-off-by: Christian König <christian.koenig@amd.com>
Reviewed-by: Dave Airlie <airlied@redhat.com>
Link: https://patchwork.freedesktop.org/patch/389457/
That is not used any more.
v2: keep the NULL checks in TTM.
v3: remove unused variable
Signed-off-by: Christian König <christian.koenig@amd.com>
Acked-by: Daniel Vetter <daniel.vetter@ffwll.ch>
Reviewed-by: Ben Skeggs <bskeggs@redhat.com>
Link: https://patchwork.freedesktop.org/patch/388646/
We are trying to remove the io_lru handling and depend
on zero init base, offset and addr here.
v2: init addr as well
Signed-off-by: Christian König <christian.koenig@amd.com>
Reviewed-by: Ben Skeggs <bskeggs@redhat.com>
Link: https://patchwork.freedesktop.org/patch/388642/
drm-next reverted the changes to ttm_tt_create() to do the
NULL check inside the function, but drm-misc-next adds new
users of this approach.
Re-apply the NULL check change inside the function to fix this.
Signed-off-by: Christian König <christian.koenig@amd.com>
Reviewed-by: Dave Airlie <airlied@redhat.com>
Link: https://patchwork.freedesktop.org/patch/386628/
Backmerging drm-next into drm-misc-next for nouveau and panel updates.
Resolves a conflict between ttm and nouveau, where struct ttm_mem_res got
renamed to struct ttm_resource.
Signed-off-by: Thomas Zimmermann <tzimmermann@suse.de>
The names get/put are associated with reference counting
in the Linux kernel, use alloc/free instead.
Signed-off-by: Christian König <christian.koenig@amd.com>
Reviewed-by: Dave Airlie <airlied@redhat.com>
Link: https://patchwork.freedesktop.org/patch/384340/?series=80346&rev=1
This name better reflects what the object does. I didn't rename
all the pointers it seemed too messy.
Signed-off-by: Dave Airlie <airlied@redhat.com>
Acked-by: Christian König <christian.koenig@amd.com>
Reviewed-by: Ben Skeggs <bskeggs@redhat.com>
Signed-off-by: Dave Airlie <airlied@redhat.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20200804025632.3868079-60-airlied@gmail.com
under driver control, this flag isn't needed anymore,
remove the API that used to access it, and consoldiate
with the used api.
Reviewed-by: Christian König <christian.koenig@amd.com>
Reviewed-by: Ben Skeggs <bskeggs@redhat.com>
Signed-off-by: Dave Airlie <airlied@redhat.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20200804025632.3868079-56-airlied@gmail.com
The driver now controls these, the core just controls the system
memory one.
v2: init sysman explicitly and assign it as a driver manager
to simplify the lookup sequence.
Reviewed-by: Christian König <christian.koenig@amd.com>
Reviewed-by: Ben Skeggs <bskeggs@redhat.com>
Signed-off-by: Dave Airlie <airlied@redhat.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20200804025632.3868079-55-airlied@gmail.com
Drivers should all be cleaning up their memory managers
themselves now, so let the core just clean the system one up.
Remove the legacy cleaning interface.
Reviewed-by: Christian König <christian.koenig@amd.com>
Reviewed-by: Ben Skeggs <bskeggs@redhat.com>
Signed-off-by: Dave Airlie <airlied@redhat.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20200804025632.3868079-36-airlied@gmail.com
The disable path is just temporary for now, it will be dropped once has_type
is gone in a later patch.
v2: add docs.
rename to ttm_mem_type_manager namespace
Reviewed-by: Christian König <christian.koenig@amd.com>
Reviewed-by: Ben Skeggs <bskeggs@redhat.com>
Signed-off-by: Dave Airlie <airlied@redhat.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20200804025632.3868079-26-airlied@gmail.com
This will allow the driver to control the ordering here better.
Eventually the old path will be removed.
v2: add docs for new APIs.
rename new path to ttm_mem_type_manager_init/set_used(for now)
Reviewed-by: Christian König <christian.koenig@amd.com>
Signed-off-by: Dave Airlie <airlied@redhat.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20200804025632.3868079-14-airlied@gmail.com
As suggested on review, just export the memory type debug for
drivers to use, while also making the debug callback optional
(don't need to test for system as it won't init it).
rename it to be more consistent with object name for now.
(we may rename all the objects later.)
Reviewed-by: Christian König <christian.koenig@amd.com>
Reviewed-by: Ben Skeggs <bskeggs@redhat.com>
Signed-off-by: Dave Airlie <airlied@redhat.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20200804025632.3868079-9-airlied@gmail.com
It is a very strange concept to call a function which just
calls back the caller for the functions parameters.
Signed-off-by: Christian König <christian.koenig@amd.com>
Reviewed-by: Thomas Zimmermann <tzimmermann@suse.de>
Reviewed-by: Alex Deucher <alexander.deucher@amd.com>
Link: https://patchwork.freedesktop.org/patch/382085/
Instead use a boolean field in the memory manager structure.
Also invert the meaning of the field since the use of a TT
structure is the special case here.
v2: cleanup zero init.
Signed-off-by: Christian König <christian.koenig@amd.com>
Reviewed-by: Daniel Vetter <daniel.vetter@ffwll.ch>
Reviewed-by: Thomas Zimmermann <tzimmermann@suse.de>
Reviewed-by: Alex Deucher <alexander.deucher@amd.com>
Link: https://patchwork.freedesktop.org/patch/382079/
Instead of repeating that in each driver.
v2: keep the caching limitation for VMWGFX for now.
Signed-off-by: Christian König <christian.koenig@amd.com>
Reviewed-by: Daniel Vetter <daniel.vetter@ffwll.ch>
Reviewed-by: Alex Deucher <alexander.deucher@amd.com>
Reviewed-by: Thomas Zimmermann <tzimmermann@suse.de>
Link: https://patchwork.freedesktop.org/patch/382078/
Just drop the argument from this.
This does ask the question if this is the function vmwgfx
should be using or should it be doing an evict all like
the other drivers.
Reviewed-by: Christian König <christian.koenig@amd.com>
Reviewed-by: Zack Rusin <zackr@vmware.com>
Signed-off-by: Dave Airlie <airlied@redhat.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20200728034254.20114-1-airlied@gmail.com
We can't pipeline that during eviction because the memory needs
to be available immediately.
v2: fix how we cleanup the BOs resources
Signed-off-by: Christian König <christian.koenig@amd.com>
Reviewed-by: Felix Kuehling <Felix.Kuehling@amd.com>
Tested-by: Alex Sierra <alex.sierra@amd.com>
Link: https://patchwork.freedesktop.org/patch/379039/
-----BEGIN PGP SIGNATURE-----
iQFSBAABCAA8FiEEq68RxlopcLEwq+PEeb4+QwBBGIYFAl8UzA4eHHRvcnZhbGRz
QGxpbnV4LWZvdW5kYXRpb24ub3JnAAoJEHm+PkMAQRiGQ7cH/3v+Gv+SmHJCvaT2
CSu0+7okVnYbY3UTb3hykk7/aOqb6284KjxR03r0CWFzsEsZVhC5pvvruASSiMQg
Pi04sLqv6CsGLHd1n+pl4AUYEaxq6k4KS3uU3HHSWxrahDDApQoRUx2F8lpOxyj8
RiwnoO60IMPA7IFJqzcZuFqsgdxqiiYvnzT461KX8Mrw6fyMXeR2KAj2NwMX8dZN
At21Sf8+LSoh6q2HnugfiUd/jR10XbfxIIx2lXgIinb15GXgWydEQVrDJ7cUV7ix
Jd0S+dtOtp+lWtFHDoyjjqqsMV7+G8i/rFNZoxSkyZqsUTaKzaR6JD3moSyoYZgG
0+eXO4A=
=9EpR
-----END PGP SIGNATURE-----
Merge v5.8-rc6 into drm-next
I've got a silent conflict + two trees based on fixes to merge.
Fixes a silent merge with amdgpu
Signed-off-by: Dave Airlie <airlied@redhat.com>
The original intention was to avoid CPU page table unmaps
when BOs move between the GTT and SYSTEM domain.
The problem is that this never correctly handled changes
in the caching attributes or backing pages.
Just drop this for now and simply unmap the CPU page
tables in all cases.
Signed-off-by: Christian König <christian.koenig@amd.com>
Reviewed-by: Daniel Vetter <daniel.vetter@ffwll.ch>
Link: https://patchwork.freedesktop.org/patch/378240/
Just use the use_io_reserve_lru flag. It doesn't make much
sense to have two flags.
Signed-off-by: Christian König <christian.koenig@amd.com>
Reviewed-by: Daniel Vetter <daniel.vetter@ffwll.ch>
Link: https://patchwork.freedesktop.org/patch/378238/
Stop touching the backend private pointer alltogether and
make sure we never put the same mem twice by.
Signed-off-by: Christian König <christian.koenig@amd.com>
Reviewed-by: Madhav Chauhan <madhav.chauhan@amd.com>
Link: https://patchwork.freedesktop.org/patch/375613/
We only need the page array when the BO is about to be accessed.
So not only populate, but also create it on demand.
v2: move NULL check into ttm_tt_create()
v3: fix the occurrence in ttm_bo_kmap_ttm as well
Signed-off-by: Christian König <christian.koenig@amd.com>
Reviewed-by: Michael J. Ruhl <michael.j.ruhl@intel.com>
Link: https://patchwork.freedesktop.org/patch/373182/
Instead of signaling failure by setting the node pointer to
NULL do so by returning -ENOSPC.
v2: add memset() to make sure that mem is always initialized.
v3: drop memset() only set mm_node = NULL, move mm_node init in amdgpu
Signed-off-by: Christian König <christian.koenig@amd.com>
Reviewed-by: Michael J. Ruhl <michael.j.ruhl@intel.com>
Link: https://patchwork.freedesktop.org/patch/373181/
GPU address handling is device specific and should be handle by its device
driver.
Signed-off-by: Nirmoy Das <nirmoy.das@amd.com>
Reviewed-by: Christian König <christian.koenig@amd.com>
Link: https://patchwork.freedesktop.org/patch/372937/
Signed-off-by: Christian König <christian.koenig@amd.com>
-----BEGIN PGP SIGNATURE-----
iQFSBAABCAA8FiEEq68RxlopcLEwq+PEeb4+QwBBGIYFAl7mfkAeHHRvcnZhbGRz
QGxpbnV4LWZvdW5kYXRpb24ub3JnAAoJEHm+PkMAQRiGTG4H/1iH1Psd32XKoO63
XuOashbqnsYLCKckbFg0RgHIDTp6d0wypF5m3dSYqpifWUaEf8SkTGnm0geV9zc1
axDUdaqr+OvhDPZiSLt4cFu2M5yvGb4/WR76qjKzxWd+LNUGOVpx1GvFXl/5wPdp
/lYR3mqZqNffwK2QuZ1m8X1gy5fzr4esoyZDK71dP88uBlFmbQLJsYiCRfQ7GDsm
r574gz7SKjPoPFPj7qJ3CLsnOruZiL36uxmYZWC2hnkNaFzLfMFiE7OqrEZK/YV6
Gc8ZZGtN+Otr+UbojagjTjRM6uy3qWBYEMX4GcRt9M2jS8LGKsYcxV75h0U5qlyz
m8QAA0o=
=jy/I
-----END PGP SIGNATURE-----
Merge v5.8-rc1 into drm-misc-fixes
Beginning a new release cycles for what will become v5.8. Updating
drm-misc-fixes accordingly.
Signed-off-by: Thomas Zimmermann <tzimmermann@suse.de>
ttm_bo_add_move_fence() invokes dma_fence_get(), which returns a
reference of the specified dma_fence object to "fence" with increased
refcnt.
When ttm_bo_add_move_fence() returns, local variable "fence" becomes
invalid, so the refcount should be decreased to keep refcount balanced.
The reference counting issue happens in one exception handling path of
ttm_bo_add_move_fence(). When no_wait_gpu flag is equals to true, the
function forgets to decrease the refcnt increased by dma_fence_get(),
causing a refcnt leak.
Fix this issue by calling dma_fence_put() when no_wait_gpu flag is
equals to true.
Signed-off-by: Xiyu Yang <xiyuyang19@fudan.edu.cn>
Signed-off-by: Xin Tan <tanxin.ctf@gmail.com>
Reviewed-by: Christian König <christian.koenig@amd.com>
Link: https://patchwork.freedesktop.org/patch/370221/
Signed-off-by: Christian König <christian.koenig@amd.com>
Calling ttm_bo_cleanup_memtype_use() destroys the TT object
which in turn could result in warnings without this.
Signed-off-by: Christian König <christian.koenig@amd.com>
Acked-by: Daniel Vetter <daniel.vetter@ffwll.ch>
Acked-by: Sumit Semwal <sumit.semwal@linaro.org>
Link: https://patchwork.freedesktop.org/patch/359288
The assert sometimes incorrectly triggers when pinned BOs are destroyed.
Signed-off-by: Christian König <christian.koenig@amd.com>
Reviewed-by: Huang Rui <ray.huang@amd.com>
Tested-by: Pierre-Eric Pelloux-Prayer <pierre-eric.pelloux-prayer@amd.com>
Link: https://patchwork.freedesktop.org/patch/356737/
Pipeline removal of the BOs backing store when no placement is given
during validation.
Signed-off-by: Christian König <christian.koenig@amd.com>
Acked-by: Daniel Vetter <daniel.vetter@ffwll.ch>
Link: https://patchwork.freedesktop.org/patch/353994/?series=73646&rev=1
This allows release_notify to add and remove fences from deleted objects.
Signed-off-by: Christian König <christian.koenig@amd.com>
Reviewed-by: xinhui pan <xinhui.pan@amd.com>
Link: https://patchwork.freedesktop.org/patch/352750/
When non-imported BOs are resurrected for delayed delete we replace
the dma_resv object to allow for easy reclaiming of the resources.
v2: move that to ttm_bo_individualize_resv
v3: add a comment to explain what's going on
Signed-off-by: Christian König <christian.koenig@amd.com>
Reviewed-by: xinhui pan <xinhui.pan@amd.com>
Reviewed-by: Daniel Vetter <daniel.vetter@ffwll.ch>
Link: https://patchwork.freedesktop.org/patch/352927/
